All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Network admins/sysadmins, need a bit of help
So we purchased a cpanel hosting via one of the hosts- not from LET, and they have their servers with fiberstate. We got a pretty good node, ryzen 9950x so everything should be snappy for our woocommerce website.
The website did work great for a couple of days, and then I started noticing decline in the performance. Support said, everything is fine at their end, however our clarity data also showed decline in several factors including TTFB. I suspected that it was a network issue, so I ran MTR and the below are the results. PS: No updates or any sort of changes were made on the website during this duration.
**Day 1 - when the site was okaish
**
Host Loss% Snt Last Avg Best Wrst StDev
1. 192.168.29.1 0.0% 55 182.7 15.0 1.3 223.6 46.6
2. 10.35.184.1 1.9% 55 136.5 26.6 2.6 247.3 59.2
3. 172.16.3.136 0.0% 54 138.7 60.2 12.4 246.1 77.3
4. 192.168.161.84 1.9% 54 94.8 67.1 11.0 231.1 75.6
192.168.161.82
5. 192.168.232.133 1.9% 54 48.2 60.2 11.4 228.7 66.1
6. 192.168.232.163 1.9% 54 12.4 69.0 11.6 230.5 74.3
7. 192.168.144.52 5.6% 54 13.8 72.5 12.1 233.5 77.1
192.168.144.54
8. (waiting for reply)
9. (waiting for reply)
10. (waiting for reply)
11. 103.198.140.176 0.0% 54 126.0 90.2 38.6 256.5 68.3
12. 103.198.140.213 0.0% 54 229.0 218.7 147.7 397.7 70.8
13. 103.198.140.213 0.0% 54 180.9 201.0 147.6 370.4 59.7
14. hurricane.mrs.franceix.net 96.2% 54 146.0 189.5 146.0 233.0 61.5
15. ve952.core1.bio1.he.net 75.5% 54 235.4 214.6 156.0 366.1 63.5
16. 100ge0-17.core2.orf2.he.net 51.9% 54 244.4 285.7 244.0 409.1 53.7
17. port-channel24.core3.ash1.he.net 88.7% 54 241.3 282.5 241.3 443.5 79.9
18. port-channel11.core2.ash1.he.net 98.1% 54 249.9 249.9 249.9 249.9 0.0
19. port-channel2.core3.mci3.he.net 69.8% 54 257.3 271.4 257.0 327.9 23.0
20. port-channel8.core2.den1.he.net 73.6% 54 274.6 292.9 273.7 345.6 25.3
21. port-channel10.core2.slc1.he.net 62.3% 54 467.4 347.3 280.1 489.0 89.3
22. fiberstate.e0-2.switch1.slc4.he. 1.9% 54 315.9 337.1 278.2 510.7 69.5
port-channel10.core2.slc1.he.net
23. ae0-dist.slc01.fiberstate.com 0.0% 54 316.8 336.1 274.9 531.6 65.9
fiberstate.e0-2.switch1.slc4.he.net
24. gull.linuxhostingserver.com 0.0% 54 281.9 344.9 281.7 524.4 71.4
ae0-dist.slc01.fiberstate.com
**Day 2- I noticed significant improvement in the loading times, without any changes on the website so I ran the tests again. Below is the data for day 2
**
Host Loss% Snt Last Avg Best Wrst StDev
1. 192.168.29.1 0.0% 14 1.5 1.5 1.2 1.6 0.1
2. 10.35.80.1 0.0% 14 3.8 4.3 3.2 5.6 0.8
3. 172.16.3.134 0.0% 14 9.6 15.3 7.6 83.8 20.1
4. 192.168.161.84 0.0% 14 7.9 12.0 7.9 41.6 8.7
192.168.161.82
5. 192.168.232.137 0.0% 14 8.8 11.3 6.6 45.3 10.0
6. 192.168.232.162 21.4% 14 8.2 8.6 7.6 10.1 0.9
7. 192.168.144.52 7.1% 14 7.2 20.9 7.2 108.6 31.8
192.168.144.54
8. (waiting for reply)
9. (waiting for reply)
10. (waiting for reply)
11. 103.198.140.174 7.7% 13 36.5 41.6 35.0 90.1 15.5
12. 103.198.140.27 0.0% 13 140.9 144.5 136.2 174.3 12.3
13. 103.198.140.27 0.0% 13 135.7 137.3 135.4 142.6 2.0
14. (waiting for reply)
15. ve952.core1.bio1.he.net 66.7% 13 146.7 147.2 146.7 148.3 0.7
16. 100ge0-17.core2.orf2.he.net 83.3% 13 231.1 228.6 226.1 231.1 3.5
17. port-channel24.core3.ash1.he.net 7.7% 13 263.6 264.1 263.1 266.3 1.0
**Day 3- I noticed significant decline in performance again today, and ran these tests again, here's todays data.
**
The site is super slow today again, after being better yesterday. I ran mtr and here are the details
Host Loss% Snt Last Avg Best Wrst StDev
1. 192.168.29.1 0.7% 2712 53.6 28.6 1.1 271.8 53.7
2. 10.35.80.1 78.6% 2712 4.8 34.1 2.2 271.6 55.9
3. 172.16.3.134 78.6% 2712 8.0 38.6 7.0 255.9 53.6
4. 192.168.161.82 78.7% 2712 7.9 40.3 7.5 269.6 56.8
192.168.161.84
5. 192.168.232.137 79.0% 2712 7.9 36.3 6.0 248.1 54.6
6. 192.168.232.163 81.1% 2712 6.7 40.6 6.5 288.4 57.9
7. 192.168.144.52 81.4% 2712 9.1 44.3 6.7 307.7 60.3
192.168.144.54
8. (waiting for reply)
9. (waiting for reply)
10. (waiting for reply)
11. 103.198.140.174 81.3% 2712 43.0 77.6 34.1 262.4 65.1
12. 103.198.140.27 80.6% 2712 352.3 205.9 142.5 438.2 72.9
13. 103.198.140.27 80.6% 2712 309.2 204.3 142.5 455.2 71.1
14. hurricane.mrs.franceix.net 98.4% 2712 165.5 214.4 142.6 385.4 76.0
15. ve952.core1.bio1.he.net 96.0% 2712 169.4 217.7 154.1 381.1 74.6
16. 100ge0-17.core2.orf2.he.net 97.3% 2712 230.8 293.1 230.7 457.7 68.5
17. port-channel24.core3.ash1.he.net 97.7% 2712 302.9 292.3 224.1 449.2 69.8
18. port-channel11.core2.ash1.he.net 99.5% 2712 245.1 298.1 242.3 433.0 61.5
19. port-channel2.core3.mci3.he.net 92.9% 2712 253.0 308.6 252.2 538.9 63.2
20. port-channel8.core2.den1.he.net 96.2% 2712 279.4 323.4 256.3 480.0 69.5
21. port-channel10.core2.slc1.he.net 91.4% 2712 267.0 335.0 265.3 607.3 69.8
22. fiberstate.e0-2.switch1.slc4.he. 78.5% 2711 275.7 336.3 274.1 616.2 68.6
23. ae0-dist.slc01.fiberstate.com 78.5% 2711 276.2 335.4 273.7 614.5 68.2
24. gull.linuxhostingserver.com 77.8% 2711 354.2 337.5 276.7 581.7 67.0
I am really not a network expert, however, I fed these into chat gpt, and it confirmed there was indeed some network congestion going on. Part of Chat GPT reply below:
From this latest MTR output, it’s clear the network path has become unstable again—explaining why your site feels slow. Here’s what stands out in the data:
- Large Jumps in Latency (High “Wrst” Values)
Hop 4 jumps from a low “Best” latency (8 ms) to a “Worst” of 226 ms.
Hop 5 has a Worst of 248 ms, Hop 6 up to 230 ms, Hop 7 up to 237 ms.
This indicates highly inconsistent latency as you move past your ISP’s local network. The large difference between “Best” and “Worst” times means the route is experiencing spikes or congestion.
I would really appreciate some pointers in this regard since the hosts wants me to send other things including clarity data, etc, however, I have already shared the MTR data with them.
Is there a packet loss/congestion actually happening based on the above data?
Comments
Try opening the BBR?
We need to see the full end-to-end mtr to know for sure. Packet loss and jitter in the middle of the path is inconclusive. Why did you buy this hosting package? From that first mtr, you would have been much happier with a host pretty much anywhere else on the planet. I'm not sure you could find a higher RTT if you tried.
Apologies! Here you go: this is the today's data.
Host Loss% Snt Last Avg Best Wrst StDev
1. 192.168.29.1 0.7% 2712 53.6 28.6 1.1 271.8 53.7
2. 10.35.80.1 78.6% 2712 4.8 34.1 2.2 271.6 55.9
3. 172.16.3.134 78.6% 2712 8.0 38.6 7.0 255.9 53.6
4. 192.168.161.82 78.7% 2712 7.9 40.3 7.5 269.6 56.8
192.168.161.84
5. 192.168.232.137 79.0% 2712 7.9 36.3 6.0 248.1 54.6
6. 192.168.232.163 81.1% 2712 6.7 40.6 6.5 288.4 57.9
7. 192.168.144.52 81.4% 2712 9.1 44.3 6.7 307.7 60.3
192.168.144.54
8. (waiting for reply)
9. (waiting for reply)
10. (waiting for reply)
11. 103.198.140.174 81.3% 2712 43.0 77.6 34.1 262.4 65.1
12. 103.198.140.27 80.6% 2712 352.3 205.9 142.5 438.2 72.9
13. 103.198.140.27 80.6% 2712 309.2 204.3 142.5 455.2 71.1
14. hurricane.mrs.franceix.net 98.4% 2712 165.5 214.4 142.6 385.4 76.0
15. ve952.core1.bio1.he.net 96.0% 2712 169.4 217.7 154.1 381.1 74.6
16. 100ge0-17.core2.orf2.he.net 97.3% 2712 230.8 293.1 230.7 457.7 68.5
17. port-channel24.core3.ash1.he.net 97.7% 2712 302.9 292.3 224.1 449.2 69.8
18. port-channel11.core2.ash1.he.net 99.5% 2712 245.1 298.1 242.3 433.0 61.5
19. port-channel2.core3.mci3.he.net 92.9% 2712 253.0 308.6 252.2 538.9 63.2
20. port-channel8.core2.den1.he.net 96.2% 2712 279.4 323.4 256.3 480.0 69.5
21. port-channel10.core2.slc1.he.net 91.4% 2712 267.0 335.0 265.3 607.3 69.8
22. fiberstate.e0-2.switch1.slc4.he. 78.5% 2711 275.7 336.3 274.1 616.2 68.6
23. ae0-dist.slc01.fiberstate.com 78.5% 2711 276.2 335.4 273.7 614.5 68.2
24. gull.linuxhostingserver.com 77.8% 2711 354.2 337.5 276.7 581.7 67.0
What kind of loss is that😂
This network is messed up honestly.
You have severe packet loss on your local network which correlates the performance. Check it outside of your own network.
The packet loss starts from hop 2 and is consistent the whole way after that, so this problem is on your end, not the host's end.